INDEX

Explanations

text related to specific sections or segments within a larger context

oai_token-act-pair · gpt-3.5-turbo

references to sections or divisions in a document or text

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

 accompl

-0.72

 score

-0.69

 forests

-0.69

 Whis

-0.69

 clos

-0.69

 placebo

-0.69

sat

-0.68

 outl

-0.68

 whis

-0.68

 veto

-0.67

POSITIVE LOGITS

section

4.18

sections

3.48

sectional

1.81

sect

1.74

paragraph

1.67

Section

1.51

chapter

1.41

article

1.22

terms

1.18

slice

1.14

Activations Density 0.024%

text related to specific sections or segments within a larger context

references to sections or divisions in a document or text

No Comments

No Known Activations