INDEX

Explanations

phrases related to observations or insights

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate observation or witnessing events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

nor

-0.77

ranged

-0.76

toe

-0.74

orth

-0.72

raft

-0.68

save

-0.67

anium

-0.66

ãĥĦ

-0.66

phe

-0.66

Ãª

-0.65

POSITIVE LOGITS

 examples

1.16

 parallels

1.14

 similarities

1.13

 firsthand

1.10

 glimps

1.09

 instances

1.04

 hints

1.03

 signs

1.02

 flashes

1.00

how

0.98

Activations Density 0.142%

phrases related to observations or insights

phrases that indicate observation or witnessing events

No Comments

No Known Activations