INDEX

Explanations

phrases related to making significant changes or improvements

oai_token-act-pair · gpt-3.5-turbo

terms related to improvement and revitalization processes

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

Chart

-0.66

\">

-0.64

ANY

-0.62

 Attend

-0.62

PLIED

-0.61

Mess

-0.61

ILA

-0.60

True

-0.59

 Writ

-0.59

 LIMITED

-0.59

POSITIVE LOGITS

issance

0.95

 havoc

0.77

 rethink

0.76

xual

0.75

afort

0.74

ulate

0.74

undown

0.72

igger

0.72

foundland

0.69

naissance

0.67

Activations Density 0.154%

phrases related to making significant changes or improvements

terms related to improvement and revitalization processes

No Comments

No Known Activations