INDEX

Explanations

verbs related to expressing opinions or points of view

oai_token-act-pair · gpt-3.5-turbo

phrases emphasizing assertion or claims

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

ItemTracker

-0.81

 teasp

-0.65

 subur

-0.65

srf

-0.63

Viz

-0.62

 Unloaded

-0.62

 Sabha

-0.61

 Going

-0.60

 Beats

-0.59

 turf

-0.58

POSITIVE LOGITS

enance

1.03

uate

1.01

ISE

0.74

ulate

0.73

 aloud

0.71

lessly

0.68

 RELEASE

0.67

 oneself

0.65

leep

0.65

dress

0.65

Activations Density 0.316%

verbs related to expressing opinions or points of view

phrases emphasizing assertion or claims

No Comments

No Known Activations