INDEX

Explanations

references to legal situations and consequences

oai_token-act-pair · gpt-3.5-turbo

references to legal consequences or penalties

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ciating

-0.70

intel

-0.66

 lately

-0.66

âĢ¢âĢ¢âĢ¢âĢ¢

-0.64

laz

-0.63

 crappy

-0.62

ĸļ

-0.60

 anymore

-0.57

ãĤ·ãĥ£

-0.57

 ignorance

-0.56

POSITIVE LOGITS

 likely

0.97

 could

0.93

 risks

0.86

could

0.85

 chances

0.85

 hopes

0.84

likely

0.81

would

0.81

 potentially

0.81

 vows

0.81

Activations Density 0.236%

references to legal situations and consequences

references to legal consequences or penalties

No Comments

No Known Activations