INDEX

Explanations

references to editing or editing-related activities

oai_token-act-pair · gpt-3.5-turbo

sections that denote edits or modifications in textual content

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

userc

-0.79

gren

-0.75

etsy

-0.73

milo

-0.71

hene

-0.71

mable

-0.67

imens

-0.67

iru

-0.67

ocket

-0.66

ongyang

-0.64

POSITIVE LOGITS

 edit

0.86

 Blizzard

0.73

 Wikipedia

0.70

...]

0.69

][

0.69

].

0.68

âĶĢâĶĢ

0.68

]).

0.66

0.64

ËĪ

0.64

Activations Density 0.019%

references to editing or editing-related activities

sections that denote edits or modifications in textual content

No Comments

No Known Activations