INDEX

Explanations

phrases related to occurrences or accidents involving physical harm or danger

oai_token-act-pair · gpt-3.5-turbo

keywords indicating actions or important events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

 carbohyd

-0.96

 conflic

-0.91

 inertia

-0.91

 parity

-0.90

 satell

-0.90

 disinteg

-0.89

 disarm

-0.89

 attrition

-0.89

 bonded

-0.87

 superiority

-0.87

POSITIVE LOGITS

 About

1.59

 Again

1.54

 With

1.52

 During

1.46

 Without

1.45

 Before

1.45

 Into

1.43

 Than

1.43

 Anyway

1.42

 Through

1.42

Activations Density 0.351%

phrases related to occurrences or accidents involving physical harm or danger

keywords indicating actions or important events

No Comments

No Known Activations