INDEX

Explanations

words related to strong negative emotions or actions

oai_token-act-pair · gpt-3.5-turbo

negative connotations or feelings associated with behavior or conditions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 Petersen

-0.79

ota

-0.74

 Passage

-0.72

 Fitzgerald

-0.71

wana

-0.69

 Achievements

-0.68

inav

-0.68

 Tanz

-0.67

 Procedures

-0.67

 Shades

-0.66

POSITIVE LOGITS

ous

1.11

ful

1.05

ously

1.01

fully

0.99

 afore

0.91

lessly

0.87

istically

0.85

osal

0.81

shire

0.80

istic

0.80

Activations Density 0.028%

words related to strong negative emotions or actions

negative connotations or feelings associated with behavior or conditions

No Comments

No Known Activations