INDEX

Explanations

social media actions, such as following, blocking, or unblocking, and their associated dates

oai_token-act-pair · gpt-3.5-turbo

references to social media interactions or updates

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

 corrid

-0.74

ierrez

-0.67

accompanied

-0.63

 pregn

-0.63

istani

-0.61

renheit

-0.60

keley

-0.59

 incor

-0.59

 appell

-0.58

sterdam

-0.58

POSITIVE LOGITS

 Comments

0.84

 Posts

0.83

 Allows

0.83

 Adds

0.83

 Latest

0.81

 Availability

0.81

 Invalid

0.78

 Likes

0.77

 Tweet

0.76

0.74

Activations Density 0.118%

social media actions, such as following, blocking, or unblocking, and their associated dates

references to social media interactions or updates

No Comments

No Known Activations