INDEX

Explanations

phrases indicating a method or course of action

oai_token-act-pair · gpt-3.5-turbo

phrases indicating methods or approaches to achieve specific outcomes

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

ĸļ

-0.95

usters

-0.81

asts

-0.72

riks

-0.70

encer

-0.69

anmar

-0.65

etheus

-0.64

uster

-0.64

oubted

-0.63

aredevil

-0.63

POSITIVE LOGITS

ward

0.89

finding

0.86

fare

0.84

point

0.80

allo

0.71

way

0.71

 forward

0.70

forward

0.69

NE

0.69

bm

0.69

Activations Density 0.025%

phrases indicating a method or course of action

phrases indicating methods or approaches to achieve specific outcomes

No Comments

No Known Activations