INDEX

Explanations

technical terms related to hardware components and software tools

oai_token-act-pair · gpt-3.5-turbo

technical specifications and functionality descriptions

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

politics

-0.65

 paradox

-0.61

 perceptions

-0.60

 recession

-0.59

 perception

-0.58

Political

-0.57

 ridicule

-0.57

 negativity

-0.56

angs

-0.56

angering

-0.55

POSITIVE LOGITS

 handy

0.82

 spons

0.79

 optionally

0.77

 stocked

0.72

 nifty

0.70

 suffice

0.69

 optional

0.69

 conveniently

0.68

 assorted

0.68

optional

0.67

Activations Density 1.139%

technical terms related to hardware components and software tools

technical specifications and functionality descriptions

No Comments

No Known Activations