INDEX

Explanations

references to political endorsements or statements of support

oai_token-act-pair · gpt-3.5-turbo

references to endorsements in the context of political support

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

eno

-0.76

 nerv

-0.76

fare

-0.75

bang

-0.71

istor

-0.69

nel

-0.68

ĸļ

-0.68

awar

-0.68

istics

-0.67

rooms

-0.66

POSITIVE LOGITS

 endorsements

1.15

 endorsement

1.10

 endorsing

1.09

 endors

1.00

 candidate

0.99

 endorse

0.98

 endorsed

0.95

 candidacy

0.93

 loyalty

0.85

 candidates

0.84

Activations Density 0.055%

references to political endorsements or statements of support

references to endorsements in the context of political support

No Comments

No Known Activations