INDEX

Explanations

admiration or positive regard towards something or someone

oai_token-act-pair · gpt-3.5-turbo

instances of the word "admire" in different contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

escal

-0.68

inventoryQuantity

-0.68

plan

-0.66

matter

-0.65

ammy

-0.65

 pregn

-0.65

onew

-0.65

unch

-0.62

budget

-0.62

TPPStreamerBot

-0.61

POSITIVE LOGITS

 admire

1.17

 admired

1.00

 admiration

0.90

rities

0.82

 BaseType

0.76

sburg

0.74

dogs

0.71

gdala

0.69

lihood

0.69

ESCO

0.69

Activations Density 0.016%

admiration or positive regard towards something or someone

instances of the word "admire" in different contexts

No Comments

No Known Activations