INDEX

Explanations

words related to instructional or informational guides

oai_token-act-pair · gpt-3.5-turbo

titles of guides and analytical reports

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

 forfe

-0.67

bucks

-0.67

 cancell

-0.67

OOOO

-0.66

umers

-0.65

 tether

-0.65

etsu

-0.63

 isEnabled

-0.63

 hook

-0.63

 chops

-0.63

POSITIVE LOGITS

 Retrieved

1.05

 Edited

0.84

 Historical

0.82

0.74

 Archive

0.73

 Contemporary

0.72

 edited

0.70

rift

0.70

0.69

Activations Density 0.727%

words related to instructional or informational guides

titles of guides and analytical reports

No Comments

No Known Activations