INDEX

Explanations

social media posts or updates

oai_token-act-pair · gpt-3.5-turbo

references to images or pictures

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

cffff

-0.69

edient

-0.69

Ļ½

-0.65

quartered

-0.65

NetMessage

-0.63

ifted

-0.60

schild

-0.60

 clust

-0.59

 Leilan

-0.59

 contracting

-0.58

POSITIVE LOGITS

colo

1.01

pic

0.90

://

0.85

amera

0.84

twitter

0.84

pic

0.84

chrom

0.81

 pics

0.77

img

0.76

apixel

0.76

Activations Density 0.009%

social media posts or updates

references to images or pictures

No Comments

No Known Activations