INDEX

Explanations

phrases related to processes or events that involve change or advancement

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate occurrences or actions

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

iege

-0.70

ridge

-0.64

 Madness

-0.63

amiya

-0.61

velt

-0.61

 Britann

-0.60

ively

-0.58

 Vide

-0.57

 Archdemon

-0.57

 depiction

-0.56

POSITIVE LOGITS

rid

1.15

tin

0.89

TING

0.84

cloneembedreportprint

0.81

ãĥ³ãĤ¸

0.79

 progressively

0.79

href

0.75

terson

0.74

 bumped

0.73

bog

0.72

Activations Density 0.123%

phrases related to processes or events that involve change or advancement

phrases that indicate occurrences or actions

No Comments

No Known Activations