INDEX

Explanations

phrases related to leaving or departing

oai_token-act-pair · gpt-3.5-turbo

phrases related to escape or evasion

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

urally

-0.81

umbing

-0.78

matically

-0.75

 underestimate

-0.74

 challeng

-0.70

orescent

-0.69

mos

-0.69

Îº

-0.68

immer

-0.68

 reactive

-0.68

POSITIVE LOGITS

away

0.90

aways

0.83

 Opportun

0.80

 Wildcats

0.77

heid

0.77

lane

0.73

0.69

vantage

0.68

AU

0.68

tery

0.67

Activations Density 0.017%

phrases related to leaving or departing

phrases related to escape or evasion

No Comments

No Known Activations