INDEX

Explanations

occurrences of the number "two" in different contexts

oai_token-act-pair · gpt-3.5-turbo

instances of the word "two"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ugu

-0.76

asta

-0.72

awaru

-0.72

amaru

-0.69

ubs

-0.69

Fed

-0.67

rir

-0.66

ysical

-0.66

aukee

-0.63

uffer

-0.63

POSITIVE LOGITS

 thirds

1.42

 dozen

1.00

 halves

0.98

 weeks

0.95

 hundred

0.88

teen

0.88

een

0.88

fold

0.86

thirds

0.86

 decades

0.81

Activations Density 0.075%

occurrences of the number "two" in different contexts

instances of the word "two"

No Comments

No Known Activations