INDEX

Explanations

phrases related to various roles or positions in different contexts

oai_token-act-pair · gpt-3.5-turbo

discussions about the influence and role of social media in various contexts

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

IDA

-0.81

INESS

-0.71

OTAL

-0.65

reports

-0.64

pedia

-0.64

BUS

-0.63

olf

-0.62

!)

-0.62

++;

-0.60

 Instructor

-0.60

POSITIVE LOGITS

 outweigh

0.86

 overshadow

0.73

clus

0.69

 favour

0.67

 prevail

0.64

ĸļ

0.63

 influencing

0.63

 accessibility

0.62

 civilisation

0.61

theless

0.60

Activations Density 0.394%

phrases related to various roles or positions in different contexts

discussions about the influence and role of social media in various contexts

No Comments

No Known Activations