INDEX

Explanations

words related to ranks, positions, or levels

oai_token-act-pair · gpt-3.5-turbo

various forms of the word "final" in different contexts

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

-0.80

berman

-0.76

è¦ļéĨĴ

-0.74

PLIED

-0.73

schild

-0.72

ullivan

-0.72

awaru

-0.72

fortune

-0.71

Beg

-0.69

EEK

-0.67

POSITIVE LOGITS

ysis

1.33

inal

1.11

ity

1.05

ient

0.92

pha

0.90

idad

0.89

phrine

0.85

culus

0.84

ITY

0.82

tarian

0.82

Activations Density 0.016%

words related to ranks, positions, or levels

various forms of the word "final" in different contexts

No Comments

No Known Activations