INDEX

Explanations

terms related to locations or events

oai_token-act-pair · gpt-3.5-turbo

references to numerical values, particularly in the context of events or objects

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

ospons

-0.77

etts

-0.65

ayson

-0.62

essel

-0.62

xus

-0.60

 Petersen

-0.60

 Pair

-0.59

 Lerner

-0.58

ihar

-0.58

ilon

-0.58

POSITIVE LOGITS

FTWARE

0.79

-'

0.77

'?

0.62

urai

0.61

Alert

0.61

 Grav

0.61

wagon

0.61

iverse

0.60

 nostalgia

0.60

 Dynam

0.60

Activations Density 0.128%

terms related to locations or events

references to numerical values, particularly in the context of events or objects

No Comments

No Known Activations