INDEX

Explanations

mentions of personnel actions and controversies in official or political contexts

oai_token-act-pair · gpt-3.5-turbo

patterns of authority and accountability in governmental or organizational contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.9.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ankind

-0.80

 Compatibility

-0.72

Fest

-0.71

Population

-0.68

population

-0.68

 Reincarn

-0.67

 ç¥ŀ

-0.67

Interstitial

-0.67

species

-0.65

 cubic

-0.65

POSITIVE LOGITS

 briefed

1.20

 testifying

1.15

 advising

1.13

 subpoen

1.13

 leaked

1.10

 resign

1.09

 resigned

1.08

 lobbying

1.07

 testified

1.07

 lobbyist

1.04

Activations Density 0.845%

mentions of personnel actions and controversies in official or political contexts

patterns of authority and accountability in governmental or organizational contexts

No Comments

No Known Activations