INDEX

Explanations

terms related to the "use" of various tools, strategies, or technologies

oai_token-act-pair · gpt-3.5-turbo

references to the concept of "use" in various 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

 Mous

-0.72

Bom

-0.72

 Tale

-0.67

 Barg

-0.65

 Bert

-0.65

 Gord

-0.64

 delic

-0.64

 Alger

-0.64

 notch

-0.63

 suspic

-0.63

POSITIVE LOGITS

fulness

1.13

fully

1.12

FUL

0.96

ful

0.92

full

0.90

ability

0.86

onomy

0.84

ivation

0.82

iences

0.82

rences

0.80

Activations Density 0.063%

terms related to the "use" of various tools, strategies, or technologies

references to the concept of "use" in various contexts

No Comments

No Known Activations