INDEX

Explanations

instances where blame is assigned to someone or something

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "blamed" and variations of it

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

marked

-0.86

edom

-0.81

cre

-0.76

shr

-0.75

mark

-0.73

visors

-0.72

intend

-0.72

mental

-0.70

inational

-0.70

ser

-0.69

POSITIVE LOGITS

 blame

0.84

 blames

0.77

 blaming

0.76

 scapego

0.73

 blamed

0.71

 lash

0.70

 lashes

0.70

 Gateway

0.67

ifully

0.66

PsyNetMessage

0.63

Activations Density 0.020%

instances where blame is assigned to someone or something

occurrences of the word "blamed" and variations of it

No Comments

No Known Activations