INDEX

Explanations

one-on-one interactions or activities

oai_token-act-pair · gpt-3.5-turbo

references to a series of sequential actions or processes

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

agher

-0.66

ONSORED

-0.59

ynt

-0.57

Ø§ÙĦ

-0.57

ãģ®ç

-0.56

 Sind

-0.56

Nib

-0.54

packing

-0.53

 wedd

-0.53

ãĤ©

-0.53

POSITIVE LOGITS

die

0.77

apest

0.69

etime

0.67

etheless

0.62

itability

0.61

«

0.61

pecting

0.59

elfth

0.59

ĥ

0.59

itiveness

0.58

Activations Density 0.065%

one-on-one interactions or activities

references to a series of sequential actions or processes

No Comments

No Known Activations