INDEX

Explanations

phrases indicating progress or improvement

oai_token-act-pair · gpt-3.5-turbo

phrases indicating a long-term impact or consequence

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

ubs

-0.61

item

-0.58

ousse

-0.57

odor

-0.56

 clen

-0.56

urities

-0.53

ãĤ¦ãĤ¹

-0.53

 THEN

-0.53

Ã©e

-0.52

ulsion

-0.52

POSITIVE LOGITS

 route

0.95

 step

0.81

 mile

0.79

Course

0.71

 Distance

0.71

 distance

0.71

 Step

0.69

 ways

0.69

 routes

0.65

 furthe

0.65

Activations Density 0.062%

phrases indicating progress or improvement

phrases indicating a long-term impact or consequence

No Comments

No Known Activations