INDEX

Explanations

phrases or words related to things being oriented or described in a backward manner

oai_token-act-pair · gpt-3.5-turbo

the word 'backwards' and related concepts of reversing or retracing steps

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

raltar

-0.87

ateurs

-0.81

rament

-0.79

"},"

-0.74

riz

-0.73

atum

-0.70

aukee

-0.70

AIN

-0.70

rien

-0.69

========

-0.69

POSITIVE LOGITS

wards

0.97

ward

0.95

stairs

0.90

 compat

0.87

step

0.79

gradation

0.78

 compatibility

0.76

dash

0.74

 backward

0.70

graded

0.69

Activations Density 0.013%

phrases or words related to things being oriented or described in a backward manner

the word 'backwards' and related concepts of reversing or retracing steps

No Comments

No Known Activations