INDEX

Explanations

headings or titles within text, indicated by the character sequence ĊĊ

oai_token-act-pair · gpt-3.5-turbo

numerical data or statistical references

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

boro

-0.73

 verbally

-0.67

 oath

-0.66

 Holo

-0.62

 Ghostbusters

-0.62

 offended

-0.62

 cough

-0.62

gery

-0.60

 exerc

-0.60

 vocals

-0.59

POSITIVE LOGITS

Newsletter

1.19

Indeed

1.11

Nevertheless

1.06

Nonetheless

1.05

Still

1.02

Story

0.97

Related

0.92

Incre

0.90

ccording

0.90

Moreover

0.90

Activations Density 0.467%

headings or titles within text, indicated by the character sequence ĊĊ

numerical data or statistical references

No Comments

No Known Activations