INDEX

Explanations

verbs or adjectives related to causing harm or negative impact

oai_token-act-pair · gpt-3.5-turbo

words related to negative impacts or harm inflicted on individuals or groups

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

USS

-0.77

 NAACP

-0.57

 boarded

-0.56

ipe

-0.56

 shit

-0.54

ussen

-0.52

icipated

-0.51

POSE

-0.51

 Gree

-0.50

aughed

-0.49

POSITIVE LOGITS

by

1.33

 therein

1.09

 herein

1.00

 during

0.94

BY

0.94

by

0.94

during

0.82

By

0.81

 thereto

0.80

 upon

0.79

Activations Density 0.181%

verbs or adjectives related to causing harm or negative impact

words related to negative impacts or harm inflicted on individuals or groups

No Comments

No Known Activations