INDEX

Explanations

the term "tu" with varying levels of importance assigned to it

oai_token-act-pair · gpt-3.5-turbo

references to tuples and related data structures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

 Angels

-0.71

dark

-0.70

å°Ĩ

-0.70

 DRAG

-0.68

 Philipp

-0.68

older

-0.66

 Seraph

-0.65

cycles

-0.65

 Siren

-0.65

Saf

-0.64

POSITIVE LOGITS

tu

1.08

ework

0.95

ples

0.93

fts

0.93

ition

0.93

uple

0.92

pling

0.90

Tu

0.88

pped

0.84

ileaks

0.81

Activations Density 0.010%

the term "tu" with varying levels of importance assigned to it

references to tuples and related data structures

No Comments

No Known Activations