INDEX

Explanations

phrases referring to judgment or evaluation of actions or beliefs

oai_token-act-pair · gpt-3.5-turbo

references to individuals and their actions or beliefs

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Apocalypse

-0.64

 unbeat

-0.63

anmar

-0.63

interstitial

-0.60

urers

-0.60

catentry

-0.59

 Recovery

-0.58

 Sisters

-0.56

 sweeping

-0.56

orc

-0.55

POSITIVE LOGITS

ever

1.03

 knows

1.00

 cares

0.93

 remembers

0.93

 ever

0.92

soever

0.92

 thinks

0.90

 opposes

0.89

 invests

0.87

 participates

0.87

Activations Density 0.098%

phrases referring to judgment or evaluation of actions or beliefs

references to individuals and their actions or beliefs

No Comments

No Known Activations